xlm-roberta-large-TASTESet-ner

This model is a fine-tuned version of xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4970
• Precision: 0.8662
• Recall: 0.8989
• F1: 0.8822
• Accuracy: 0.8889

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	31	1.8592	0.3077	0.4305	0.3589	0.4376
No log	2.0	62	1.3188	0.4793	0.5445	0.5098	0.5884
No log	3.0	93	1.1581	0.5382	0.6134	0.5733	0.6391
No log	4.0	124	1.1373	0.6480	0.5964	0.6211	0.6522
No log	5.0	155	0.8784	0.6969	0.7370	0.7164	0.7425
No log	6.0	186	0.7242	0.7472	0.7823	0.7643	0.7930
No log	7.0	217	0.6340	0.7869	0.8258	0.8058	0.8225
No log	8.0	248	0.5766	0.7832	0.8562	0.8180	0.8391
No log	9.0	279	0.5200	0.8087	0.8702	0.8383	0.8583
No log	10.0	310	0.4981	0.8495	0.8722	0.8607	0.8642
No log	11.0	341	0.4732	0.8510	0.8836	0.8670	0.8762
No log	12.0	372	0.4884	0.8593	0.8801	0.8696	0.8746
No log	13.0	403	0.4701	0.8444	0.8893	0.8663	0.8825
No log	14.0	434	0.4759	0.8576	0.8898	0.8734	0.8814
No log	15.0	465	0.4765	0.8596	0.8945	0.8767	0.8840
No log	16.0	496	0.4817	0.8610	0.8984	0.8793	0.8881
0.7221	17.0	527	0.4904	0.8572	0.8989	0.8775	0.8869
0.7221	18.0	558	0.4971	0.8640	0.8969	0.8802	0.8869
0.7221	19.0	589	0.4954	0.8595	0.9024	0.8804	0.8894
0.7221	20.0	620	0.4970	0.8662	0.8989	0.8822	0.8889

Framework versions

• Transformers 4.26.0
• Pytorch 1.13.1+cu117
• Datasets 2.9.0
• Tokenizers 0.13.2